Toshihiko Kimura via sheffield.ac.uk Attachments29 Sep (12 days ago) to Vik Dear Prof. Dhillon, I suppose the u' filter for DECam (we provided in 2012) would be better for you than ODI u'. It was short pass type u' filter without long pass film. I'll ask our engineer to make the data with updated refractive index. I overwrote the theoretical curve of u' (short pass type) on the Excel file I sent on 12 September. In ODI u', the cut-off wavelength was 390 nm. Meanwhile in DECam, it was 400 nm. This should be your choice. Even if we do not add long pass film, the transmission is decreasing around 330 nm or shorter due to the absorption of coating material (Ta2O5). Sincerely, Toshihiko -----Original Message----- From: Toshihiko Kimura Sent: Thursday, September 15, 2016 7:10 PM To: Vik Dhillon Subject: Re: Theoretical design(High-efficiency SDSS filters) Dear Prof. Dhillon, Can this be pushed slightly further to the blue, e.g. to around 300nm? I suppose the u' filter for DECam (we provided in 2012) would be better for you than ODI u'. It was short pass type u' filter without long pass film. I'll ask our engineer to make the data with updated refractive index. Thanks, Toshihiko From: Vik Dhillon Sent: Thursday, September 15, 2016 7:03 PM To: Toshihiko Kimura Subject: Re: Theoretical design(High-efficiency SDSS filters) Hi Toshihiko, Thank you - I shall await the theoretical curves for the original SDSS filter set before making a final decision (although I am sure we will want to order these fantastic new filters!). One additional question - the blue cut-off of the u' filter is at ~330nm. Can this be pushed slightly further to the blue, e.g. to around 300nm? This will gain us even more throughput, as the atmosphere is still allowing some light through at these wavelengths. Regards, Vik. --- Prof. Vik Dhillon, Dept of Physics & Astronomy, Univ of Sheffield, Sheffield S3 7RH, UK +44 114 222 4528; www.vikdhillon.staff.shef.ac.uk --- On 15 September 2016 at 11:00, Toshihiko Kimura wrote: Dear Prof. Dhillon, I will ask our engineer to make the theoretical design of each old SDSS filter. Please give him time to do. I assume that these curves are for the coating performance only, and do not include any absorption due to the substrate? No, these curves include the absorption of coating materials. You can find it easily in u' band design (bluer side in passband). We will use fused silica as substrate. Absorption by fused silica from 300 nm to 1000 nm is almost negligible. Sincerely, Toshihiko From: Vik Dhillon Sent: Thursday, September 15, 2016 6:37 PM To: Toshihiko Kimura Subject: Re: Theoretical design(High-efficiency SDSS filters) Hi Toshihiko, Thank you for the theoretical data. I assume that these curves are for the coating performance only, and do not include any absorption due to the substrate? In order for me to make a fair comparison between these new filters and the old SDSS filters, would it be possible for you to send me theoretical curves for the old SDSS filters? I have looked through all of my emails from you and all I can find are the as-built curves, which obviously include the effects of the coloured glass and manufacturing errors. It would be better to compare the theoretical curves of both the new and old SDSS filters, if possible. Regards, Vik. --- Prof. Vik Dhillon, Dept of Physics & Astronomy, Univ of Sheffield, Sheffield S3 7RH, UK +44 114 222 4528; www.vikdhillon.staff.shef.ac.uk --- On 12 September 2016 at 11:04, Toshihiko Kimura wrote: Dear Prof. Dhillon, I would like to send the theoretical curve of each "high-efficiency" SDSS filters for HiPERCAM. Our engineer used the HSC filter design as a reference for g' ~ z', and ODI u' filter design as a reference for HiPERCAM u'. The design is done on the basis of AOI 0 degree. I understand that each cut-on / off wavelength should be optimized in consideration of actual HiPERCAM dichroic curve and f/2.45 beam finally. Of course, each filter must be parfocal. [HiPERCAM g' ~ z' filter] Cut-on and off wavelength tolerance: each +/-5 nm T(average) > 95% or over in passband T > 90% in passband (i.e. valley of ripple must not be less than 90%) T (average) < 0.01% 300 nm - 1100 nm T < 0.1% 300 nm - 1100 nm [HiPERCAM u' filter] Cut-on wavelength tolerance: +4 nm, -6 nm Cut-on wavelength tolerance: +/-4 nm T > 85% or over in passband T (average) < 0.01% 300 nm - 1000 nm T < 0.1% 300 nm - 1100 nm Though I don't specify the number of slopes in passband, the actual curve will be close to the theoretical curve. Sincerely, Toshihiko From: Toshihiko Kimura Sent: Wednesday, September 07, 2016 6:10 PM To: Vik Dhillon Subject: Re: ROM estimation(High-efficiency SDSS filters) Dear Prof. Dhillon, I will be able to get back to you on 8 ~ 9 September. Please wait for a moment, our engineer is going to finalize the theoretical design. Thanks, Toshihiko From: Vik Dhillon Sent: Thursday, August 25, 2016 7:24 PM To: Toshihiko Kimura Subject: Re: ROM estimation(High-efficiency SDSS filters) Hi Toshihiko, Thanks for the information - looking forward to seeing the throughput curves! Regards, Vik. On 25 Aug 2016 09:32, "Toshihiko Kimura" wrote: Dear Prof. Dhillon, you mean that the biggest improvements will be made in the u' and g' filters - and the r' i' z' filters won't be very different to the ones we currently have. Is that right? Sorry that my message was not clear. I would like to clarify again. ---------------------------------------------- It is easier to make the monolithic r', i', z' filters than u' and g'. It means it is not difficult for us to improve and achieve the high transmission in r', i', z' passband. Meanwhile, it is very hard for any vendors to make the monolithic u' filter. It means it is not easy to achieve the high transmission in passband. It is more difficult to make the monolithic g' filter than r', i', z' filters, but easier than u'. The relation is like, (hard) u' >>> g' > r', i' > z' (easy). ---------------------------------------------- As you would notice, the transmission of conventional u' filter is lowest among SDSS u' ~ z' filters. It is caused by the original filter prescription, colored glass such as UG11 or KG3. Though I said, "it is not easy to achieve the high transmission in passband" as above, we achieved > 90% transmission in e.g. ODI, DECam, MegaCam (CFHT), or PAUCam monolithic u filter. Compared to the conventional u' filter (low transmission, not sharp slopes due to UG11), you can expect it will be much improved. On the other hand, the conventional r', i', z' filters already had > 90% transmission in passband, though colored glass (OG550 or RG695) was used and eventually bluer slopes in passband were not sharp. Like HSC r filter, the transmission will be much improved. However the difference between conventional and high-efficiency is not so big as u'. ---------------------------------------------- it would be really useful if we could compare the transmission curves of the new filters you have quoted for with our current set of SDSS filters for HiPERCAM. Would you be able to send me the curves? Absolutely, we should do that. I was worried that you would lose your interest due to the price, but now I find you did not. I asked our engineer to make each theoretical design and think about our suggested specifications. I will be able to get back to you on 8 ~ 9 September. Sincerely, Toshihiko From: Vik Dhillon Sent: Wednesday, August 24, 2016 6:03 PM To: Toshihiko Kimura Subject: Re: ROM estimation(High-efficiency SDSS filters) Hi Toshihiko, Thank you for the quotation. I am definitely interested in procuring these filters! I have a few questions: - If I understand you correctly, you mean that the biggest improvements will be made in the u' and g' filters - and the r' i' z' filters won't be very different to the ones we currently have. Is that right? - In this case, I wonder if it is worth us procuring just the u' and g' filters and forgetting about the r' i' z' filters, since we won't see much improvement in these. What do you think? - To help make our decision, it would be really useful if we could compare the transmission curves of the new filters you have quoted for with our current set of SDSS filters for HiPERCAM. Would you be able to send me the curves? Regards, Vik. --- Prof. Vik Dhillon, Dept of Physics & Astronomy, Univ of Sheffield, Sheffield S3 7RH, UK +44 114 222 4528; www.vikdhillon.staff.shef.ac.uk --- On 22 August 2016 at 09:13, Toshihiko Kimura wrote: Dear Prof. Dhillon, I would like to apologize my slow response. I'm afraid you would feel ridiculous for our estimation. Indeed, it is much more expensive than the current SDSS filters for HiPERCAM. However, if you are still interested in changing the SDSS filters with "conventional recipes" (i.e. colored glass), we are willing to discuss about the specifications. I fully understand that without the significant improvement, it is not worth for you to replace them. since we are going to be changing the filter bandpass shapes so much (i.e. trying to get as close to 100% top-hat transmission curves as possible), It is hardest for any vendors to improve the transmission of u-band filter. Only the solution is definitely to make the monolithic u filter, but it drives the difficulty of coating significantly up. The transmission of u will be much improved, however it is not easy to fulfill "almost 100% top-hat transmission". (Please refer to the actual transmission curve of u-band filter for ODI at WIYN Observatory). http://www.asahi-spectra.com/opticalfilters/large_filters.html The transmission of g-band will be also improved (especially bluer wavelength in passband). r, i and z-band will be close to the actual transmission of HSC r filter. (Please refer to the same link as above) These filters are completely custom products. So the specifications of spectral response should be discussed in consideration of the actual transmission of HiPERCAM dichroics and f/2.47 beam. All the filters should be made with the monolithic 5.43 mm fused silica substrate. Sincerely, Toshihiko From: Vik Dhillon Sent: Tuesday, July 26, 2016 7:15 PM To: Toshihiko Kimura Subject: Re: High-efficiency SDSS filters Hi Toshihiko, Thank you for your reply. A rough estimate of the cost would be extremely useful at this stage - thank you for agreeing to provide this. Regards, Vik. --- Prof. Vik Dhillon, Dept of Physics & Astronomy, Univ of Sheffield, Sheffield S3 7RH, UK +44 114 222 4528; www.vikdhillon.staff.shef.ac.uk --- On 26 July 2016 at 03:03, Toshihiko Kimura wrote: Dear Prof. Dhillon, My response is also slow, I don't offer the quote yet. I noticed that "high efficiency" was important performance for some astronomers. By the conversation with you in Edinburgh, I decided to prepare Sloan filter set (or UBVRI) with "high-efficiency version". In your case, you already purchased standard SDSS filter set (PT) from us. If you deicide to replace, I understand that there must be strong motivation to do. That is, such filters must be significantly improved in terms of band shape and transmission. Of course, wavelengths must be optimized with HiPERCAM dichroics (i.e. completely customized filters) I suppose it would take some time for each other to agree the final specifications (if you decide to buy). Apart from the final specifications, I think I should send the rough estimation, because cost is important information for your management. I need to prepare the bid proposal for University of Tokyo by next week, but I will do the effort to submit it soon. Sincerely, Toshihiko From: Vik Dhillon Sent: Sunday, July 24, 2016 6:50 PM To: Toshihiko Kimura Subject: Re: High-efficiency SDSS filters Hi Toshihiko, Thanks for your email, and sorry for my slow response. I understand that these will not be cheap filters! One thing I would like to say is that, since we are going to be changing the filter bandpass shapes so much (i.e. trying to get as close to 100% top-hat transmission curves as possible), the details of whether the filters obey the PT or SDSS 2.5m system is not so critical. What is critical is that the bandpass edges of each filter are identical to the HiPERCAM filters you already made for us (so the PT system), as only then will they be compatible with the cut points of the HiPERCAM dichroics! I hope this is clear. Have you had a chance to look any more into this with your engineers yet? Regards, Vik. --- Prof. Vik Dhillon, Dept of Physics & Astronomy, Univ of Sheffield, Sheffield S3 7RH, UK +44 114 222 4528; www.vikdhillon.staff.shef.ac.uk --- On 14 July 2016 at 09:57, Toshihiko Kimura wrote: Dear Prof. Dhillon, Sorry for the delay in my response. I talked with our engineer about high-efficiency Sloan filters. why is sputtering better than IAD? There are three technical aspects. One of the technical reasons is that it is more advantageous to control each film-thickness as designed compared to IAD. Another reason is repeatability of production by sputtering is much better than IAD. Last reason is that surface roughness coated by sputtering is (slightly) better than IAD. By the way, we provided SDSS filters for ULTRACAM in 2001. Under the suggestion by Prof. Mamoru Doi (University of Tokyo), we manufactured SDSS filters on the basis of PT system at that time. (Please see the e-mail copy from him to you as below.) SDSS filters for ULTRASPEC and HiPERCAM are compatible with ULTRACAM, i.e. PT system. However, Prof. Doi suggested to me that Asahi should provide SDSS filters on the basis of 2.5 m response several years ago. Therefore, we usually do not produce SDSS filters on the basis of PT system. What I would like to say is that it is not easy for us to regard SDSS filters (PT system) as standard products. It is related to the price. Anyway, I would like to discuss with our engineer more. SDSS filters coated by sputtering are definitely better than IAD in terms of high-efficiency. I would like to think more how we can offer "reasonable" price in consideration of improved performance. Sincerely, Toshihiko -----Original Message----- From: Mamoru Doi Sent: Monday, July 23, 2001 10:00 PM To: vik.dhillon@shef.ac.uk Cc: t-kimura@asahi-spectra.co.jp ; Takashi Ichikawa Subject: SDSS filters Dear Dr. Vikram Dhillon, Mr.Kimura, Asahi spectra Co, asked me to explain the reason why the SDSS band response has changed from the original. The main reason is that interference filters of the 2.5-m imager changed their band responses (in g', r', and i'). We put filters into dewars, i.e. into vacuo. Then water component in the interference surfaces was removed, and refractive index changed. We didn't expect this, but we can not replace the filters. The imager is very complicated and the filters are deep inside. (Ref. "The Sloan Digital Sky Survey Photometric Camera" J.E. Gunn et al. AJ, 116, 3040 (1998) ) The filters Asahi spectra made this time have the band responses which are close to those of PT, the 20-inchi Photometric Telescope at the APO. The primary standard stars to determine the zero points of the 2.5-m photometry were observed with USNO 40-inch telescope system which has the original band responses. The PT connects the USNO 40-inch photometry and the 2.5-m phtometry, and PT system has the band responses which are almost in the "middle" of the USNO responses and the 2.5-m responses. I hope this e-mail answers your question. Sincerely, -Mamoru Mamoru Doi Institute of Astronomy School of Science Univ. of Tokyo voice +81-422-34-5084 fax. +81-422-34-5041 doi@ioa.s.u-tokyo.ac.jp From: Vik Dhillon Sent: Wednesday, July 06, 2016 8:34 PM To: Toshihiko Kimura Subject: Re: High-efficiency SDSS filters Hi Toshihiko, Thank you for your reply. I'm really pleased to hear that you are interested in working on these. I understand that these will be more expensive than the normal coloured glass SDSS filters. Could I ask a question? - why is sputtering better than IAD? Also, if IAD is cheaper, do you think it would be possible to make these top-hat SDSS-like filters from IAD? Anyway, I'm looking forward to hearing what your engineers say and receiving a quote. Regards, Vik. --- Prof. Vik Dhillon, Dept of Physics & Astronomy, Univ of Sheffield, Sheffield S3 7RH, UK +44 114 222 4528; www.vikdhillon.staff.shef.ac.uk --- On 6 July 2016 at 08:20, Toshihiko Kimura wrote: Dear Prof. Dhillon, Thank you for stopping at our booth in Edinburgh. I was asked about HiPERCAM dichroic mirrors by another UK astronomer. I'm proud that Asahi had an opportunity to provide these dichroics to the unique, state-of-the-art instrument. Your comment, higher-efficiency "SDSS-like" filters, was very interesting to me. I understood that like UBVRcIc, "classical" SDSS filter curve was important for astronomers. In fact, the astronomer from Carnegie Observatories said the same thing as you. but I would obviously like them all to be top-hat functions with near 100% efficiency at their peak. I understand you need the transmission curve like HSC r filter we exhibited. It was coated by magnetron sputtering technology. When we provided broad bandpass filters for e.g. Binospec or PAUCam, we fabricated them with monolithic fused silica substrate (i.e. without colored glass) by ion assisted deposition (IAD) technology. Transmission in passband was drastically improved. But apart from the production cost, sputtering method is better than IAD. I'll go to our factory tomorrow because Japanese astronomers will visit us tomorrow and also Friday. I would like to talk with our engineer about your inquiry. I can say that we will be able to produce what you want, but my concern is the cost. Best regards, Toshihiko -----Original Message----- From: Vik Dhillon Sent: Tuesday, July 05, 2016 10:14 PM To: Toshihiko Kimura Subject: High-efficiency SDSS filters Hi Toshihiko, It was a pleasure to meet you at SPIE last week. I was wondering if we could carry on our discussion about the higher-efficiency "SDSS-like" filters for HiPERCAM? I am very interested in ordering a set. Because we have already bought our dichroics for HiPERCAM, I would need these new filters to have the same cut points as the original SDSS filters, but I would obviously like them all to be top-hat functions with near 100% efficiency at their peak. Do you think you could provide me with a quotation for such a set of filters for HiPERCAM (so the same dimensions and optical thickness as you delivered for the SDSS filters for HiPERCAM)? Regards, Vik. --- Prof. Vik Dhillon, Dept of Physics & Astronomy, Univ of Sheffield, Sheffield S3 7RH, UK +44 114 222 4528; www.vikdhillon.staff.shef.ac.uk --- Attachment: Revised theoretical des ... u'g'r'i'z' filter).xls